Examples

Sequential Benchmark

The example assumes that you want to run a benchmark that shall be started using simple bash scripts. All the following instruction assume that the current working directory is the root directory of the benchmark-tool project. To begin, the two executables clasp-3.4.0 and runlim-2.0.0rc12 have to be copied (or symlinked) into the ./programs folder.
Now, run:
$ bgen ./runscripts/runscript-seq.xml
This creates a set of start scripts in the ./output folder.
To start the benchmark, run:
$ ./output/clasp-big/houat/start.py
Once the benchmark is finished, run:
$ beval ./runscripts/runscript-seq.xml | bconv -o result.ods
Finally, open the file:
$ soffice result.ods

Cluster Benchmark

This example assumes that you want to run a benchmark on a cluster, i.g. on the HPC cluster at the university of Potsdam. Again, all the following instruction assume that the current working directory is the root directory of the benchmark-tool project. Once again make sure, the two executables clasp-3.4.0 and runlim-2.0.0rc12 have been copied (or symlinked) into the ./programs folder.
Now, run:
$ bgen ./runscripts/runscript-dist.xml
This creates a set of start scripts in the ./output folder.
To start the benchmark, run (on the cluster):
$ ./output/clasp-one-as/hpc/start.sh
Once the benchmark is finished, run:
$ beval ./runscripts/runscript-dist.xml | bconv -o result.ods
Finally, open the file:
$ soffice result.ods

Runscripts

This tool comes with a collection of example runscripts to help you get started.

While runscript-example.xml gives a small example on how basic sequential and cluster benchmarks can be defined. runscript-seq.xml and runscript-dist.xml show more possibilities. runscript-all tries to be a most complete example runscript.

Examples for the encoding support feature can be found here.

For a more detailed explanation of a runsript and its components check here